Diederik Stapel, who was a star in his field as a professor of social psychology and a welcome guest in television programs, was fired by Tilburg University in 2011 after it became known that he had committed fraud on a large scale. For example, he claimed to conduct research in high schools, but in reality he fabricated research data himself.
That led to mediagenic results. For example, the thought of large pieces of meat would make people selfish. A lot of rubbish on the street would make people racist. Stapel’s inventions have been cited at least 115 times in scientific publications in the past five years, according to research by de Volkskrant.
Occasionally his work was referred to as an example of fraud, but in most cases this was done in all seriousness. Scientists referred to Stapel’s cheating work as support for a certain theory, or wrote that they would use the same approach.
The fact that fake research continues to circulate in this way is harmful, says Lex Bouter emeritus professor of methodology and integrity at VU University Amsterdam. ‘Incorrect findings and conclusions live on in this way. It pollutes scientific literature.’
When asked, several scientists who have quoted from Stapel’s fraud investigations in recent years say that they did not know that the article in question had been withdrawn, for example because they came across an old version online, or because they had read the article in question when it was not yet known that Stapel had committed fraud.
Most of them find it annoying that they have made a mistake, but point out that they do not elaborate on his work in terms of content, for example because the theory that Stapel ‘proved’ has actually been substantiated by other scientists. ‘Fortunately, we did not base our research on his findings,’ says British marketing professor Yogesh Dwivedi. According to him, researchers should be more careful, but publishers of trade journals ‘could also do more’.
‘Ghost Investigations’
‘Ghost studies’, which have been formally withdrawn by journals and yet continue to appear in the latest professional literature, are a persistent problem in science, says Professor Bouter. ‘Publications that have been officially withdrawn are cited less often, but sometimes not much less.’ These are unnecessary errors because there is good software that checks whether manuscripts refer to retracted articles, says Bouter.
In the case of Stapel, the damage is limited to ‘scientific noise’, says Bouter. But these kinds of mistakes can cause great damage, even beyond science. The professor of integrity points to the fraud case of Rotterdam internist Don Poldermans, who cheated with research into beta blockers in heart operations. His findings were so positive that the drugs became standard prescriptions in hospitals across Europe for many years.
Bouter: ‘It later turned out that those studies were flawed and that the medicine did not dramatically reduce mortality, but actually increased it. The guidelines for prescribing beta-blockers were not changed until three years after the fraud came out – and the article was withdrawn – resulting in many patients dying prematurely.’
It is ten years ago on Monday that the final report on the Stapel fraud was published. The shocking conclusion of ‘failing science’ was not only that Stapel was able to commit fraud for a long time before he was caught, but also that many scientists systematically cheated with figures in order to arrive at better results.
Diederik Stapel does not want to answer questions from de Volkskrant.
Saturday in de Volkskrant: ten years after the devastating report on the Stapel fraud: has science improved? Already online to read.